economics Nature Is Weird

Amazon product ratings dropped globally during the pandemic because customers were simply in a bad mood.

April 20, 2026

Original Paper

How Do Large-Scale Negative Events Bias Online Review Systems? Implications for Decision Support

SSRN · 6595202

The Takeaway

Online review systems are highly sensitive to large scale societal shocks. When people are stressed or isolated, they give lower ratings to everything they buy. A product that received five stars in a stable year might only get three stars during a pandemic. We treat consumer reviews as objective data about the quality of a toaster or a pair of shoes. These ratings are actually a reflection of the collective emotional state of the world. A bad review today might just be a symptom of a customer having a bad week.

From the abstract

Online review systems are core decision support systems (DSS) that guide consumers, firms, and algorithms. Their effectiveness depends on the integrity of review information, which large negative events can systematically distort. Using the COVID-19 pandemic as an exogenous shock, we draw on affect-as-information theory to argue that incidental mood shifts bias consumer evaluations beyond actual product experiences. Combining a unique dataset of Amazon reviews with a Difference-in-Differences de